An analysis of import expansion policies.

Dinopoulos, Elias ; Kreinin, Mordechai E.

AN ANALYSIS OF IMPORT EXPANSION POLICIES

1. INTRODUCTION

Policy makers appear to display unlimited ingenuity in devising new and ever more complex measures to influence international trade. No sooner did the profession come to understand the economics of trade-restricting devices, such as tariffs, quotas and voluntary export restraints (VERs) (the discirminatory nature of VERs has been recently noted and analyzed in Dinopoulos and Kreinin [1988; 1989]), then a new policy instrument was introduced, christened by Bhagwati [1987] as Voluntary Import Expansion (VIE).

In several recent instances the United States insisted that Japan "voluntarily" expand its import of certain products, such as semiconductors or beef, from the United States. While the VER restricts the volume of trade below free-market equilibrium, the VIE would expand the volume of trade beyond the equilibrium level. But both policies depart from free-trade equilibrium, and both are discriminatory in nature. A VER that limits Japanese auto exports to the U.S. favors European suppliers who are excluded from the restriction: their thems of trade improve and/or their volume of export expands. A VIE that forces the Japanese to import a minimum amount of semiconductors from the United States excludes European suppliers. A similar VIE on beef imports would discriminate against Australian and Argentinian suppliers. It constitutes a form of "export protectionism" (Bhagwati and Irwin [1987]).

How might a government administer a VIE? In the case of semiconductor imports by Japan, the possibilities are: a subsidy by the Japanese government on imports from the U.S., or an "instruction" to the large chip-using companies in Japan to purchase more American chips. (1) Alternatively, the Japanese government may buy U.S. chips for its own use, or it may favor American suppliers in awarding contracts.

In the Fall of 1987 certain Japanese steel companies agreed (under MITI pressure) to expand purchases of American coal, even though it costs $A7 per ton more than Australian coal. And in January 1988 (1) an instance of a VIE in the construction sector came to light. Japan had been under American pressure to permit U.S. construction companies to bid on large Japanese public works projects, like the Kansai airport near Osaka and the Tokyo-Bay bridge. Such projects can amount to $70 billion in the next decade. In response, the Japanese government proposed to permit joint foreign-Japanese companies to bid on these projects, and give American companies preferential treatment in forming such partnership.

This paper analyzes the economic effects of a VIE and, whenever appropriate, compares it to an export subsidy. Whereas the export subsidy is an instrument used by the exporting country, the VIE is a policy instrument employed by the importing country.

Section II presents a comparison between the two instruments within a two-country general equilibrium framework when the offer curves are elastic. Section III addresses the question of the optimum VIE and compares it to the optimum tariff. Section IV analyzes the case when the offer curve of the VIE-administering country is enelastic. Another key distinction between a VIE and an export subsidy is that the former is discriminatory between different trading partners while the latter is not. Section IV incorporates this feature by recasting the analysis in a three-country general equilibrium framework. Section V states the conclusions of the paper.

II. VIEW VERSUS EXPORT SUBSIDY: A TWO-COUNTRY ANALYSIS

(WITH ELASTIC OFFER CURVES)

In coventional trade geometry, an export subsidy (assumed, as usual, to be financed by a lump-sum tax) is shown by an outward shift (namely, an increase) of the offer curve of the subsidizing country. By contrast, a graphical representation of a VIE is more complex, and is displayed in Figure 1. Japan exports good X (which may be autos) to the U.S. and imports good Y (which may be tree) from the U.S. OJ is Japan's free-trade offer curve, and it intersects the U.S. offer curve at point A. Japan's free-trade import level is OE. Assume that under U.S. pressure Japan accepts a minimum import of OM, which is greater than OE. (3) Then the segment of Japan's free-trade offer curve below M' becomes irrelevant, and only M'J is the relevant part. In other words, if the terms of trade (rays from the origin) intersect the offer curve along segment M'J, that segment is operational. If Japan's terms of trade deteriorate to a level below the ray OM' (so that they intersect the free trade offer curve along OM'), then the minimum import takes over and Japan's offer curve becomes segment M'L. Thus, the entire offer curve is JM'L with a kink at M'. However, segment M'L does not continue indefinitely, because there is a limit to Japan's supply of its exportables.

Equilibrium Under Free Trade and a VIE:

In Figure 1, OJ and OUS are respectivel the free-trade offer curves of Japan and the U.S. yielding terms of trade OA (not drawn) and OE imports into Japan. The two trade indifference curves (TICs) tangent to ray OA at the point A, the intersection of the two offer curves, represent the welfare levels of the two countries under free trade. If minimum imports of OM are accepted by Japan, then its offer curve become JM'L with a kink at M', as explained earlier. It intersects the U.S. free-trade offer curve at point V. The new equilibrium terms of trade is ray OV, and U.S. welfare is represented by TIC [W.sup.V.sub.US.] The U.S. terms of trade improves and that Japans deteriorates relative to the free-trade position. The VIE improves U.S. welfare relative to free trade, because the U.S. has moved along its curve away from the origin, and welfare is an increasing function of the distance along a free-trade offer curve from the origin. (4)

Quantity-Equivalent Subsidy

We next define U.S. export equivalent to the VIE as the subsidy that would ensure the same quantity of U.S. exports as under the VIE, namely quantity OM. With a free-trade policy followed by Japan, its offer curve reverts to the original OM'J. The geometrical expression of the U.S. export subsidy is given by a leftward shift (increase) in the U.S. offer curve, so that it intersects the free-trade Japanese offer curve at M'. For the sake of clarity it is not drawn.

U.S. welfare under the subsidy is [W.sup.S.sub.US.] It is definitely inferior to U.S. welfare under a VIE, [(W.sup.V.sub.US)], because both the terms of trade and the volume of imports are lower.

Japan is better off with a U.S. export subsidy (at equilibrium point M') than with a VIE (at V) because its terms of trade (price of an equal quantity of Japan's imports) is higher.

Indeed the sole difference between equilibrium poins M' and V (and points in between) is the division of the rents between the two countries. If the U.S. "forces" Japan to impose a VIE, equilibrium is at V and the U.S. gets the rents. Intuitively, the VIE is equivalent to a Japanese import subsidy of the magnitude M'V, so that the subsidy is a revenue transfer to the U.S. of that amount. In the case of a U.S. export subsidy, the rent M'V accrues to Japan.

Price-Equivalent Subsidy

Under elastic offer curves the same results when equivalence is defined as the export subsidy that generates the same domestic price ratio in the U.S. as the VIE. This definition of equivalence is appropriate when the objective of the U.S. government is to maintain a given level of production in each sector. In Figure 1 the VIE-ridden equilibrium is at V, and U.S. welfare is represented by [W.sup.V.sub.US], the TIC tangent to the international and the U.S. domestic price [beta]. A price-equivalent U.S. export subsidy is one that would shift the U.S. offer curve leftward to intersect Japan's offer curve at P. Point P is defined as the intersection of Japan's free-trade offer curve (OJ) with the U.S. income-consumption path (not drawn), which is negatively sloped when both goods are normal. By the nature of the income-consumption path, the tangent to the U.S. TIC at P has the slope [beta].

U.S. welfare under the price-equivalent subsidy is represented by TIC [W.sup.S.US.] It is distinctly inferior to the VIE--ridden welfare, because by moving from V to P along the income-consumption path, we are transferring income from the U.S. to Japan. Conversely, Japan's welfare is higher with a U.S. export subsidy (at P) than with a VIE (at V).

III. THE OPTIMUM VIE (5)

Is there an optimal VIE; namely, one which would maximize U.S. welfare? In Figure 2 we extend the U.S. offer curve to its inelastic portion (beyond point B). All points of intersection between M'L and the inelastic portion of U.S. offer curve (such as point U) represent unstable equilibria. In other words, if the U.S. mandates the VIE, Japan will always choose a point on the elastic part of the American offer curve. Hence, the final outcome will never settle on the inelastic portion of the U.S. offer curve. This rules out a Japanese export whih is a Gifften good in the U.S. Likewise, if M' lies in its entirety above point B, a VIE-ridden equilibrium does not exist.

It follows that a VIE-ridden trade equilibrium can exist only on the AB portion of the U.S. offer curve. The VIE which maximizes U.S. welfare is OR; namely, the one which results in trade-equilibrium point B, where the U.S. offer curve is of unitary elasticity. The trade indifference curve [W.sup.V.sub.US] represents U.S. welfare under an optimum VIE. Thus, the optimum VIE is defined strictly in terms of the U.S. offer curve.

By contrast, the U.S. optimum tariff is defined in terms of Japan's offer curve. It results in a U.S. trade indifference curve [W.sup.t.sub.US] tangent to Japan's offer curve. It is obvious from Figure 2 that the optimum VIEW and the optimum tariff cannot be ranked.

IV. INELASTIC JAPANESE OFFER CURVE

In what follows we divide the inelastic portion of Japan's offer curve into two components: the negatively sloping part (range BC in Figure 3) and the positively sloping part (CJ), that represents the case of a U.S. export which is a Giften good in Japan. In accordance with our earlier notations, assume that OE is the free-trade imports into Japan and OM is the VIE-ridden imports. Under a VIE, Japan's offer curve would become MTCSL.

If the U.S. offer curve intersects Japan's offer curve in range BS then the analysis in Section II holds, because the VIE equilibrium would occur along SL. If the intersection occurs along ST, the VIE constraint is not binding. Finally, should the U.S. offer curve intersect Japan's free-trade offer curve in range TJ (the Giffen good range), then the VIE-ridden equilibrium is in range TM. In that latter case, the U.S. offer curve must cross Japan's offer curve from below; otherwise, the free-trade equilibrium would not be stable.

With a U.S. offer curve OUS, free-trade equilibrium is at A, and the VIE-ridden equilibrium is at V. The alternative of a U.S. subsidy cannot generate the maximum imports of OM, as the subsidy-enhanced U.S. offer curve would cut Japan's free-trade offer curve below point A. An export subsidy reduces the price of the U.S. exportable, and it being a Giffen good in Japan, the quantity demanded declines rather than rises. Thus, in order to meet the minimum import objective OM the U.S. must impose a tariff that would shift its offer curve in a way that it passes through point T.

Japan is better off under a VIE than under a U.S. tariff. To obtain the same amount of imports, it needs to export more under a tariff (MT [is greater than] MV). By similar reasoning, the U.S. is better off under a tariff.

V. VIE VERSUS EXPORT SUBSIDY IN A THREE-COUNTRY FRAMEWORK

Because the VIE is an inherently discriminatory device, while the export subsidy is not, we need to compare the two instruments in a three-country framework, which can capture discriminatory practices. We do this for one trade pattern, namely, where the U.S. imports from Europe and Japan. Analysis of alternative trade patterns yields similar results and is available from the first author upon request. The analysis is confined to elastic offer curves and to quantity equivalence.

In Figure 4 OUS is the U.S. free-trade offer curve, while OE and OJ are the offer curves of Europe (the third country) and Japan, respectively. They are added up to O(E+J) by summing the quantities traded by the two countries with the U.S. along each terms of trade line (ray from origin) (see Dinopoulos-Kreinin [1989]). The free-trade equilibrium is at F with terms of trade OF. The welfare level of the three countries is represented by the three TICs (not shown) tangent to OF at its points of intersection with the three offer curves. Japan's imports from the U.S. is OA.

We now introduce a VIE for Japan, where Japan increases its imports from OA to OM. Its offer curve becomes JM'LB, with a kink in M' (as in Section II). To derive the new (E+J) offer curve we take the following steps: (a) For all terms of trade rays steeper than OM', the (E+J) offer curve is unchanged, so the segment above point K remains intact. (b) For points below K we add up, for each terms of trade, the quantities traded by Japan along M'B (fixed imports, variable exports) with those traded by Europe along its free-trade offer curve. That yields the segment below K of the new (E+J) offer curve. Note that the segment below point L represents a reversal of Europe's trade pattern. Point L is the intersection of MB with a terms-of-trade ray representing the autarky price for Europe and Japan. The intersection of the new offer curve LK(E+J) with OUS yields the new trade equilibrium at point V. If it occurs at terms of trade steeper than OK, then the VIE constraint is not binding. If it occurs at a point below L, then the trade pattern is reversed. Hence, our interest is confined to range KL.

In our diagram, the new equilibrium is at V with terms of trade OV. The U.S. welfare is represented by [W.sup.V./.sub.US]; it is higher than its free-trade welfare. Europe's welfare declines from its free-trade position as we move down its free-trade offer curve and its terms of trade deteriorate. Japan's welfare worsens relative to free-trade (to a TIC passing through C) and its terms of trade deteriorate.

A U.S. export subsidy (non-discriminatory) which would yield the same Japanese imports from the U.S. as under the VIE (OM), would shift the U.S. offer curve to the left, such that it would pass through point K. The resulting terms of trade intersects Japan's free-trade offer curve at M', so that Japan's imports would be OM. U.S. welfare under an export subsidy is represented by the TIC [W.sup.S./.sub.US], passing through K but not tangent to OK.

In Figure 4 the U.S. terms of trade are worse under an export subsidy (OK) than under a VIE (OV). (6) Thus, the U.S. welfare is higher under a VIE. On the other hand, Japan and Europe are better off with U.S. export subsidy than a VIE. Under a U.S. subsidy, their welfare is represented by the TICs tangent to OK at points M' and I, respectively. This is the same outcome as in the two-country case.

VI. CONCLUSIONS

In a two-country world, the VIE-subsidy comparison depends on the division of the rents between the two countries. With a Japan-administered VIE the U.S. captures the rent, while under a U.S. export subsidy Japan receives the rent. Hence, the U.S. is better off with a Japanese VIE and Japan is better off with a U.S. subsidy. Introduction of an inelastic U.S. offer curve shows that the optimum VIE for the U.S. is one which results in a trade-equilibrium point where the U.S. offer curve is of unitary elasticity. It cannot be ranked relative to the optimum tariff. Stability considerations rule out a solution along the inelastic portion of the U.S. offer curve.

With an inelastic Japanese offer curve we can obtain a Giffen good in Japan. Under these circumstances no U.S. subsidy can help attain the required minimum import. Instead, the U.S. would have to levy a tariff to reach this goal. Japan is better off under a VIE than a U.S. tariff, while the reverse is true for the U.S. Introduction of a third country does not affect the welfare ranking under the two policies. Under elastic offer curves and independently of the pattern of trade, the U.S. is better off under a Japanese VIE than under a U.S. export subsidy; Japan is better off under a U.S. export subsidy than under a VIE. The third country's welfare is higher under the policy which improves its terms of trade, and that depends on the pattern of trade.

As a final point, it might be mentioned that if the VIE is designed to offset an initial Japanese trade barrier, then it is conceivable that both countries will be better off because the VIE might offset the initial distortion.

(1) See the Wall Street Journal, 21 September 1988, p. 12.

(2) See the Wall Street Journal, 13 January 1988, p. 4.

(3) It is possible that a VIE can be expressed as a percentage of the current level of imports. In this case the shape of the VIE-ridden offer curve changes but the results of the paper hold.

(4) An equivalent equilibrium would ensue if Japan introduced a trade subsidy which would shift its offer curve outward, such that it would pass through point V.

(5) We are grateful to Peter Neary for inspiring this section.

(6) This property always holds as long as the VIE constraint is binding and there are no multiple trade equilibria for the U.S.

REFERENCES

Bergsten, C. Fred, Kimberly A. Elliott, Jeffrey J. Schott and Wendy E. Takacs. Auction Quotas and United States Trade Policy, Washington, D.C.: Institute for International Economics, 1987.

Bhagwati, Jagdish N. "VERs, Quid Pro Quo DFI and VIEs: Political-Economy-Theoretic Analysis." International Economic Journal. Spring 1987, 1-14.

Bhagwati, Jagdish N. and Douglas Irwin. "The Return of the Reciprocitarians: U.S. Trade Policy Today." The World Economy, September 1987, 109-30.

Brecher, Richard A. and Jagdish N. Bhagwati. "Voluntary Export Restrictions Versus Import Restrictions: A Welfare-Theoretic Comparison," in Essays in Honour of W.M. Corden, edited by H. Kierzkowski. Basil Blackwell: Oxford, 1987,.

Dinopoulos, Elias and Mordechai E. Kreinin. "Effects of the U.S.-Japan Auto VER on European Prices and on U.S. Welfare." Review of Economics and Statistics, August 1988, 484-91.

Dinopoulos, Elias and Mordechai E. Kreinin. "Import Quotas and VERs: A Comparative Analysis in a Three-Country Framework." Journal of International Economics, February 1989, 169-78.

Kreinin, Mordechai e. and Elias Dinopoulos. "Alternative Quotas and VER Allocation Schemes--A Welfare Comparison." mimeo, University of Florida, 1988.

Meade, James E. A Geometry of International Trade. London: Allen and Irwin, 1952

Vanek, Jaroslav. International Trade: Theory and Economic Policy. Homewood, Ill.: R.D. Irwin, Inc., 1962.

During the final phases of this paper the second author was a visiting professor at Monash University, Melbourne, Australia. An earlier version of the paper was presented at the 62nd Annual Western Economic Association International Conference in Vancouver, B.C., Canada, July 1987; Monash University, Melbourne, Australia, October 1987; and the Seventh Annual Conference on International Trade Theory, University of Western Ontario, Canada, April 1988. The authors would like to thank Jagdish Bhagwati, Steven Matusz, Peter Neary, and Joe Stone for useful comments and suggests. This paper has benefited greatly from thoughtful comments and excellent suggestions by an anonymous referee.